1 / 60

Hume’s Central Principles

Hume’s Central Principles. 4. Hume on Induction. Peter Millican Hertford College, Oxford. 4(a) From the Idea of Causation to Induction. The Idea of Causation.

illias
Télécharger la présentation

Hume’s Central Principles

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Hume’s Central Principles 4. Hume on Induction Peter Millican Hertford College, Oxford

  2. 4(a)From the Ideaof Causationto Induction

  3. The Idea of Causation To understand reasoning to the unobserved (i.e. probable reasoning, though Hume has not yet used the term), “we must consider the idea of causation, and see from what origin it is deriv’d” (T 1.3.2.4). The search for the origin of this idea will shape the remainder of Treatise 1.3. There is no specific quality that characterises causes and effects, so it must be some relation between the two. (T 1.3.2.5-6) 3

  4. Contiguity and Priority We find causes and effects to be contiguous in space and time (T 1.3.2.6), though a footnote hints at a significant reservation (explored in T 1.4.5 which points out that many perceptions have no spatial location). We also find causes to be prior to their effects (T 1.3.2.7), though again Hume seems to indicate that this isn’t a particularly crucial matter (T 1.3.2.8). There still seems to be something missing … 4

  5. Necessary Connexion There follows a famous passage, which is commonly misunderstood: “Shall we then rest contented with these two relations of contiguity and succession, as affording a compleat idea of causation? By no means. An object may be contiguous and prior to another, without being consider’d as its cause. There is a NECESSARY CONNEXION to be taken into consideration; and that relation is of much greater importance, than any of the other two above-mention’d.” (T 1.3.2.11) 5

  6. To Neighbouring Fields Hume is looking for the crucial extra component (beyond single-case contiguity and succession) that makes up our idea of cause and effect It seems elusive, so he proceeds like those who “beat about all the neighbouring fields, without any certain view or design, in hopes their good fortune will at last guide them to what they search for” (T 1.3.2.13). There are two such fields … 6

  7. The Causal Maxim The first field is the Causal Maxim: “’Tis a general maxim in philosophy, that whatever begins to exist, must have a cause of existence” (T 1.3.3.1) Hume argues that this is neither intuitively nor demonstratively certain (T 1.3.3.1-8) “Since it is not from knowledge or any scientific reasoning, that we derive [this] opinion …, [it] must necessarily arise from observation and experience. … (T 1.3.3.9) 7

  8. Leading Up to Induction Treatise 1.3.4 argues that causal reasoning, if it is to result in real belief, must start from something perceived or remembered. T 1.3.5.1 sets out a corresponding agenda: “Here therefore we have three things to explain, viz. First, The original impression. Secondly, The transition to the idea of the connected cause or effect. Thirdly, The nature and qualities of that idea.” 8

  9. “Of the impressionsof the senses and memory” The title of Treatise 1.3.5 seems odd, since memory presents ideas, not impressions. But Hume’s main point here is that the perceptions of the senses and memory are alike in being more strong and lively – having more force and vivacity – than the ideas of the imagination. That force and vivacity, apparently, is what enables them to act as a “foundation of that reasoning, which we build … when we trace the relation of cause and effect” (T 1.3.5.7) 9

  10. Recap – the road to Treatise 1.3.6 Recall Hume’s aim here: He is seeking to understand our idea of necessary connexion (cf. T 1.3.2.11). This leads him to ask “Why we conclude, that … particular causes must necessarily have … particular effects, and why we form an inference from one to another?” (T 1.3.3.9). The key part of this process is “the inference from the impression to the idea” (cf. T 1.3.5.1); call this “causal inference” for short. 10

  11. 4(b)The Argument concerning Induction

  12. The Famous Argument (×3) Treatise 1.3.6 contains the famous argument concerning induction, though Hume doesn’t seem entirely to appreciate its significance – it is mainly a staging post in his search for the origin and nature of our idea of causation. In the Abstract of 1740 it is elevated to a much more prominent position, as the centre-piece of Hume’s “Chief Argument”. The fullest and clearest version is in the first Enquiry, Section 4. 12

  13. Streamlining the Argument In the Treatise, Hume’s focus is on causal inference “from the impression to the idea”. In the Abstract and Enquiry, he broadens it to ask about the foundation of “all reasonings concerning matter of fact” (A 8): “What is the nature of that evidence, which assures us of any real existence and matter of fact, beyond the present testimony of our senses, or the records of our memory.” (E 4.3) His first point is that all such [inductive] inference depends on causal relations (A 8, E 4.4). 13

  14. Causal Inference Is Not A Priori (1) In the Treatise, Hume starts from causal inference, arguing that this cannot be a priori, just because we can conceive of things coming out differently (T 1.3.6.1). Here he evinces the [common, but not obvious] assumption that any a priori inference would have to yield complete certainty. “’Tis therefore by EXPERIENCE only, that we can infer the existence of one object from that of another” (T 1.3.6.2). 14

  15. A Thought Experiment In the Abstract and Enquiry, Hume imagines Adam, newly created by God, trying to envisage the effect of a billiard-ball collision: • how could he possibly make any prediction at all in advance of experience? 15

  16. Causal Inference Is Not A Priori (2) Here also Hume’s subsequent argument is stronger, because he doesn’t rely just on conceivability, but puts more emphasis on arbitrariness: “Were any object presented to us, and were we required to pronounce concerning the effect, which will result from it, without consulting past observation; after what manner, I beseech you, must the mind proceed in this operation? It must invent or imagine some event, which it ascribes to the object as its effect; and it is plain that this invention must be entirely arbitrary. …” (E 4.9) 16

  17. Experience and Constant Conjunction The kind of experience on which causal inference is based is repeated patterns of one thing, A, followed by another, B: “Without any farther ceremony, we call the one cause and the other effect, and infer the existence of the one from that of the other.” (T 1.3.6.2) “Thus … we have … discover’d a new relation betwixt cause and effect, when we least expected it … This relation is their CONSTANT CONJUNCTION.” (T 1.3.6.3) 17

  18. “Perhaps ’twill appear in the end …” The capitalisation in T 1.3.6.3 clearly links back to T 1.3.2.11, as does the text: “Contiguity and succession are not sufficient to make us pronounce any two objects to be cause and effect, unless … these two relations are preserv’d in several instances [i.e. there’s a constant conjunction].” But how can this give rise to the new idea of necessary connexion? Anticipating T 1.3.14.20, “Perhaps ’twill appear in the end, that the necessary connexion depends on the inference, instead of the inference’s depending on the necessary connexion”. 18

  19. A Question of Faculties Since causal reasoning from [impression of] cause A to [idea of] effect B is founded on “past experience, and … remembrance of … constant conjunction” (T 1.3.6.4), “the next question is, whether experience produces the idea [of the effect B] by means of the understanding or imagination; whether we are determin’d by reason to make the transition, or by a certain association and relation of perceptions?” Hume will now argue that it can’t be reason. 19

  20. The Need for Extrapolation [All inference to matters of fact beyond what we perceive or remember seems to be based on causation, and] all our knowledge of causal relations comes from experience. Such learning from experience takes for granted that observed phenomena provide a guide to unobserved phenomena. We thus extrapolate from past to future on the assumption that they resemble. But do we have a rational basis for doing so? 20

  21. UP: The Uniformity Principle Hume then focuses on the principle (UP) presupposed by such extrapolation: “If reason determin’d us, it wou’d proceed upon that principle, that instances of which we have had no experience, must resemble those of which we have had experience, and that the course of nature continues always uniformly the same.” (T 1.3.6.4) This seems conditional: IF reason is involved, THEN it must be based on this principle. But later: “probability is founded on the presumption of a resemblance …” (T 1.3.6.7) 21

  22. UP in the Enquiry In the Enquiry is less explicitly stated: “all our experimental [experiential] conclusions proceed upon the supposition, that the future will be conformable to the past”. (E 4.19) No suggestion of conditionality (cf. also E 5.2: “in all reasonings from experience, there is a step taken by the mind” corresponding to UP). Vaguer than original Treatise UP, and so more plausible: we expect the future to “resemble” (E 4.21) the past, but not copy exactly. 22

  23. The Role of the Uniformity Principle Hume need not be suggesting that we think of UP explicitly when making inductive inferences (and T 1.3.8.13 says typically we don’t). Rather, in making an inductive inference, we manifest the assumption of UP, in basing our inferential behaviour on past experience. So inferring from past to future is ipso facto treating “the past [as a] rule for the future” (cf. E 4.21) Hence the question arises: can this assumption be founded on reason, or is there some other explanation for why we make it? 23

  24. Can UP be Founded on Argument? After stating UP in the Treatise, Hume immediately continues: “In order therefore to clear up this matter, let us consider all the arguments, upon which such a proposition may be suppos’d to be founded; and as these must be deriv’d either from knowledge or probability, let us cast our eye on each of these degrees of evidence, and see whether they afford any just conclusion of this nature.” (T 1.3.6.4) By knowledge, Hume means demonstration, as becomes evident in the next sentence. 24

  25. Both forms of argument are quickly ruled out, demonstration by the Conceivability Principle: “We can at least conceive a change in the course of nature; which … proves, that such a change is not absolutely impossible [and thus] a refutation of any pretended demonstration against it.” (T 1.3.6.5) And probable argument by circularity: “probability … is founded on the presumption of a resemblance betwixt those objects, of which we have had experience, and those, of which we have had note; and therefore ’tis impossible this presumption can arise from probability.” (T 1.3.6.7) (Hume first argues, at T 1.3.6.6-7, that probable argument is causal and hence dependent on UP.) 25

  26. Enquiry More Complete At T 1.3.6.4, Hume assumes that demon-stration and probability are the only possible foundations for UP; but in the Enquiry, he also rules out sensation and intuition: “there is no known connexion between the sensible qualities and the secret powers; and consequently, … the mind is not led to form such a conclusion concerning their constant and regular conjunction, by any thing which it knows of their nature.” (E 4.16) “The connexion … is not intuitive.” (E 4.16) 26

  27. The “Sceptical” Conclusion “even after experience has inform’d us of [causal] constant conjunction, ’tis impossible for us to satisfy ourselves by our reason, why we shou’d extend that experience beyond those particular instances, which have fallen under our observation.” (T 1.3.6.11) “even after we have experience of the operations of cause and effect, our conclusions from that experience are not founded on reasoning, or any process of the understanding” (E 4.15) “in all reasonings from experience, there is a step taken by the mind, which is not supported by any argument or process of the understanding” (E 5.2) 27

  28. Argument Summary (in 2 Slides) The essential logic of the argument can be represented using the “founded on” relation (FO), together with: p Probable inference (to the unobserved) c Causal reasoning e (Reasoning from) Experience u Uniformity Principle R Reason • d Demonstration • i Intuition • s Sensation 28

  29. Hume’s Argument concerning Induction FO(p,c) FO(p,e) FO(c,e) FO(e,u) FO(p,u) ¬FO(u,d) ¬FO(u,p) ¬FO(p,R) ¬FO(u,i) ¬FO(u,R) ¬FO(u,s) Only in Enquiry 29

  30. Four “Kinds of Evidence” (Again) So the Enquiry argument implicitly reasons: ¬FO(u,s) & ¬FO(u,i) & ¬FO(u,d) & ¬FO(u,p)  ¬FO(u,R) If UP isn’t founded on sensation, intuition, demonstration or probable inference, then it isn’t founded on Reason. Compare this passage from Hume’s Letter from a Gentleman (1745): “It is common for Philosophers to distinguish the Kinds of Evidence into intuitive, demonstrative, sensible, and moral” 30

  31. Debating Hume’s Argument A catalogue of recent interpretations: Flew 1961, Stove 1965/73: deductivism; Stroud 1977: extreme scepticism; Beauchamp & Mappes 1975, Winters 1979, B’p & Rosenberg 1981, Arnold 1983, Broughton 1983, Craig 1987, Baier 1991: refuting deductivism (hence “anti-deductivist”); Millican 1995/2002: anti-perceptual-insight; Garrett 1997: not founded on reasoning; Owen 1999: anti-stepwise-inference; Millican 2011: not founded on cognition. 31

  32. What Does “Reason” Mean? By far the most significant distinction between these interpretations is in terms of their view of “reason” or “the understanding”: Flew, Stove: deductive reasoning only Stroud: traditional “self-conscious” conception Beauchamp et al.: deductivist – but rejected Millican 1995: perceptual insight – but rejected Garrett: reason is the reasoning faculty Owen: intermediate steps – but rejected Millican 2011: reason is the cognitive faculty 32

  33. Inductive Inferences as “Reason” “… with regard to reason … The only conclusion we can draw from the existence of one thing to that of another, is by means of the relation of cause and effect …” (T 1.4.2.47) “… reason, in a strict and philosophical sense, can have an influence on our conduct … by informing us of the existence of something which is a proper object of [a passion]; or when it discovers the connexion of causes and effects, so as to afford us means of exerting any passion.” (T 3.1.1.12) 33

  34. Reducing the Field Hume is not an extreme, undiscriminating sceptic but a keen advocate of inductive science. So unless Hume is radically incon-sistent, Flew/Stove/Stroud must be wrong. Ruling out a probable foundation for UP would be otiose if “reason” were deductivist, so Beauchamp et al. must also be wrong. Millican 1995 and Owen face the objection that Hume does not apparently reject the view of reason operative in his argument. 34

  35. Agreeing with Garrett … • Don Garrett and I now agree on a fair number of points: • Hume’s “reason” is not ambiguous (a point on which he stood alone for many years); • Hume sees no obligation to prove our faculties reliable a priori (rejecting the burden of proof implied by “antecedent” scepticism – E 12.3); • The logic of his argument is incompatible with most previous interpretations (most obviously the deductivist and anti-deductivist). 35

  36. … Up to a Point • However the key disagreement remains the nature of Humean “reason”: • Garrett says “for Hume [as for Locke], reason is the faculty of reasoning: of making inferences, or providing, appreciating, and being moved by arguments.” (1997, p. 27) • I think “reason” is the overall cognitive faculty, just another word for “the understanding” or the “intellectual faculties”. • This is discussed in detail in my 2011 paper, “Hume’s ‘Scepticism’ about Induction”. 36

  37. “Reason” and “Reasoning” • We tend to hear “reasoning”, “proof” and “argument” as implying stepwise inference or ratiocination, but this is anachronistic. • Johnson in 1756 defines “argument” as: • “A reason alleged for or against any thing.” • A non-discursive sense of “reason” is: • “Argument; ground of persuasion, motive.” • And the first sense of “proof” is: • “Evidence; testimony; convincing token.” 37

  38. “Deduction” and “Ratiocination” • For stepwise inference, Johnson prefers the terms “deduction” and “ratiocination”. He gives as discursive senses of “reason”: • “The power by which man deduces one proposition from another, or proceeds from premises to consequences.” • “Ratiocination; discursive power.” • The same two terms are used for contrast when defining “intuition” and “intuitive”. 38

  39. Hume’s Usage • Hume, like Johnson, refers to “deductions” and “ratiocination” in contexts where stepwise argument is clearly intended: • T 1.3.14.2, E 5.22, M 1.4; E 4.23, E 12.17 • He also refers to “arguments” “inference” and “proof” that are “intuitive”: • T 1.3.14.35, T 2.3.2.2, E 4.21, E 8.22 n. 18 • Hume’s own theory of inductive “reasoning” implies that it is not typically stepwise! 39

  40. Hume’s Conclusion (Garrett 1997) “Hume . . . [is] making a specific claim, within cognitive psychology, about the relation between our tendency to make inductive inferences and our inferential/argumentative faculty: he is arguing that we do not adopt induction on the basis of recognising an argument for its reliability … this does not mean that inductive inferences are not themselves instances of argumentation or reasoning; … His point is rather that they are reasonings that are not themselves produced by any piece of higher level reasoning” (pp. 91-2) 40

  41. What About Intuition? • One objection to Garrett’s position (Millican 1998, p. 151) is that in the Enquiry, Hume also rules out intuition (which is not reason-ing in Garrett’s sense) as the basis of UP. • He responded in our Hume Studies debate: “Hume … in the Enquiry … expands the famous conclusion to rule out any ‘reasoning or process of the understanding,’ thereby eliminating such non-inferential processes of the understanding as intuition …” (1998, p. 184) 41

  42. Reason = The Understanding But Hume implicitly identifies reason with the understanding in many places, e.g.: “When the mind [makes an inductive inference] it is not determin’d by reason, but by certain principles, which associate together the ideas of these objects, and unite them in the imagination. Had ideas no more union in the fancy than objects seem to have to the understanding, …” (T 1.3.6.12) See also T 1.3.6.4, 1.4.1.1, 1.4.2.46, 1.4.2.57, 1.4.7.7, and compare 2.2.7.6 n. with 1.3.9.19 n. 42

  43. A Sign of Convergence… ? • Garrett has now acknowledged that Hume equates “reason” and “the understanding”: “I grant that Hume roughly interchanges the terms ‘reason’ and ‘understanding’ … The understanding generally involves the intuition of self-evident truths as well as reasoning … Peter thinks reason for Hume blew up to cover everything that the Lockean understanding did, while I think that in Hume the understanding shrank down to encompass only what reason did, plus intuition.” (2011, pp. 18-19) • To me, this shrinking is historically implausible. 43

  44. Epistemology or Cognitive Science? • Garrett’s 1997 formulation has a tension: • He sees the argument as cognitive psychology rather than epistemology: concerning the mechanism of inductive inference rather than whether or not it can be justified. • Yet he takes Hume’s conclusion to be that“we do not adopt induction on the basis of recognising an argument for its reliability, for … there is no argument … that could have this effect. … we can literally ‘give no reason’ for our making inductive inferences” (1997, p. 92). 44

  45. Hume’s Conclusion (Garrett 2002) “Hume … [is] making a specific claim, within cognitive psychology, about the underlying causal mechanism that gives rise to inductive inference: namely, that it is not itself dependent on any reasoning or inference. … this does not mean that inductive inferences are not themselves instances of argumentation or reasoning; … His point is rather that they are reasonings which are not themselves caused by any piece of reasoning (including, of course, themselves). ” (p. 333) 45

  46. Induction in General, or Individual? 46 • He also made an important clarification: “Millican understandably infers that on my interpretation ‘it is only the general practice of induction that fails to be determined by reason, and each of our particular inductive inferences is itself an instance of the operation of our reason.’ … The crucial distinction for Hume, however, is … between an inference being an instance of reasoning and the same inference being caused by (another instance of) reasoning.” (1998, pp. 180-1)

  47. The Inheritance Problem Here the problem for Garrett is to give a plausible precise account of Hume’s claim. If the claim concerns every individual inductive inference, and is a claim about the psychological mechanism involved in such inference (rather than about epistemological foundation), then it is unclear why lack of ratiocinative causation should be “inherited” by a later argument that starts from a previously-taken-as-established lemma. 47

  48. An Objection from the Logic of Hume’s Argument FO(p,c) FO(p,e) FO(c,e) FO(e,u) FO(p,u) ¬FO(u,d) ¬FO(u,p) ¬FO(p,R) ¬FO(u,i) ¬FO(u,R) ¬FO(u,s) Consider this final step 48

  49. An Implausible Non-Sequitur The final step of Hume’s argument makes no sense on Garrett’s interpretation: UP plays a role in the causation of probable inference; UP is not itself caused by a process of ratiocination; Therefore probable inference is not caused by any process of ratiocination. This is a complete non-sequitur. Probable inference could be caused by a process of ratiocination that involves UP! 49

  50. Further Logical Objections If Hume were only concerned to prove that ratiocination plays no role in the causation of induction (i.e. probable inference), then: His argument would be incomplete, because he does nothing to rule out the possibility that induction could be caused by bad argument. Much of his Enquiry argument would be redundant, because he would have no need to refute the idea that induction is founded on intuition or sensation. 50

More Related